As a followup, I guess what I'm looking for is something along the lines of what xgettext does, but directory-based instead of file- based, and also the ability to provide my own custom 'keywords' (I also use this in smarty templates, in the form of {'text'|__} ). Something like pybabel, but for php... I'm starting to wonder how complicated this should be, as it sounds like a fairly easy task...

On 9 Sep 2007, at 18:13, Constantinos Neophytou ♎ wrote:

I have a question that I'm sure has a simple solution, and if my brain worked properly these days I'd be able to find it either online or somewhere in the archives... But I can't, so here goes...

I need to parse an entire project through a regex, and produce some formatted output. The regex is quite simple, and the reason is to create a 'translation' file. I have a huge project where I've used the wordpress-style convention (which in turn is based on the gettext style convention) of translating strings, namely passing them to the __() function (this is all in php). I'm not using gettext, i've just wrapped my own code around that function (which is why i'm using __() instead of _() )

This function at the moment simply returns the input as output (since all the text strings are in english, and I haven't really translated anything yet so my translation files are empty). What I want is to parse all of my code for /__(['"](.*)['"])/ and some variations thereof, and produce a line-separated list of strings that need to be translated. I'd like to make this an automated process, because this is an ongoing project and i'd like to refresh the to-be-translated strings file frequently.

I believe poedit already does something like this, but attempting to get my head around .po and .mo files at this point just gave me a migraine (nb: if someone knows of a good, step-by-step 'for dummies' tutorial for this stuff, I'd appreciate it).

Any ideas?

Constantinos Neophytou

On 9/9/07, Constantinos Neophytou ♎ constantinos@neophytou.net wrote:

I need to parse an entire project through a regex...

Take a look at the "Grep in Project" command:

http://henrik.nyh.se/2007/06/grep-in-project-command-for-textmate

If can be easily modified to suit your needs, and it spits results on a format you can copy and paste on another document.

Also, good old commandline grep may be just what you're looking for...

On 9 Sep 2007, at 20:12, Ale Muñoz wrote:

Take a look at the "Grep in Project" command:

http://henrik.nyh.se/2007/06/grep-in-project-command-for-textmate

If can be easily modified to suit your needs, and it spits results on a format you can copy and paste on another document.

Thanks, that helped. I made a bash script that outputs the grep data, and I have the find/replace strings for TextMate's regular expression match.. What's the best way to combine the two, i.e. run the bash script, and then pipe the output in a regex find/replace, and output to a file? I'm guessing my options are either a complicated macro (don't know how to output to a file with this method, if even possible), or a Command... which I think would be preferable, but then how would I go about executing the find/replace inside the command?

Constantinos Neophytou